The nocebo response
The pill may be inactive, but the
side effects are real.
(This article was first printed in the March
2005 issue of the Harvard Mental Health
Letter. For more information or to order,
please go to http://www.health.harvard.edu/mental.)
About 20% of patients taking a sugar pill in
controlled clinical trials of a drug spontaneously
report uncomfortable side effects — an
even higher percentage if they are asked. These
effects are one kind of nocebo — a word
that means in Latin “I will harm,” as
placebo means “I will please.”
A placebo makes patients feel better for reasons
unrelated to the specific healing properties
of the treatment. A nocebo makes patients feel
worse (or does other harm) in the same way. Common
symptoms are drowsiness, headache, mild dizziness,
difficulty concentrating, and stomach upset.
Many health professionals are not aware of nocebos,
yet the reaction can cause patients to drop out
of clinical trials, stop taking drugs they need,
or end up using other drugs that complicate their
The nocebo effect can result from conditioning,
as when patients become nauseated or even vomit
on entering a room where they have recently received
chemotherapy. Medications and other treatments
take on symbolic features that can have nocebo
effects. Red is associated with stimulation,
blue with sedation, so red and blue pills may
produce those responses as unwanted side effects.
Contagious rumor is another source of nocebo
responses. Many people who have heard about penicillin
allergies, wrongly think that they are allergic
to penicillin, and report reactions.
Experiments show the potential of explicit suggestion
in medical treatment for good or ill.
- Volunteers were told that a mild electrical
current would be passed through their heads
and might cause a headache. No electrical current
was actually passed, but two-thirds of them
developed a headache.
- Patients with asthma were divided into two
groups. One was given a bronchoconstrictor,
which ordinarily makes asthma symptoms worse,
and told that it was a bronchodilator, which
normally improves the symptoms. This placebo
suggestion reduced their discomfort by nearly
50%. The second group was given a bronchodilator
and told it was a bronchoconstrictor. The nocebo
suggestion reduced the drug’s effectiveness
by nearly 50%.
- The same treatment can work as both a nocebo
and a placebo. Experimenters gave subjects
who believed they were allergic to various
foods an injection they were told contained
the allergen. It was only salt water, but it
produced allergic symptoms in many of them.
Then the experimenters injected salt water
again, this time saying it would neutralize
the effect of the previous injection — and
in many cases it did.
- An active drug has more nocebo power than
a mere sugar pill. In one study, experimental
subjects were divided into four groups. The
first was given a muscle relaxant, described
correctly; the second group was given the same
muscle relaxant but told it was a stimulant;
the third group received a sugar pill described
as a muscle relaxant, and the fourth received
the same inert pill described as a stimulant.
To no one’s surprise, subjects who thought
the pill was a stimulant were more likely to
say they felt tense. But the muscle relaxant
caused more reports of tension when described
as a stimulant than the sugar pill did. Blood
levels of the muscle relaxant were lower in
people told it was a stimulant than in those
told the truth. They may have absorbed less
of the drug because the false information activated
the sympathetic nervous system, which slows
down movements of the intestinal tract.
Anyone can experience a nocebo effect, but it
appears that the same people respond strongly
to both nocebos and placebos. In one experiment,
subjects in three groups were asked to keep a
hand in ice water as long as they could. One
group was told that this could have beneficial
effects for a period of up to five minutes (placebo
instruction). The second group was told that
it could be harmful, so the experiment would
be stopped after at most five minutes as a precaution
(nocebo instruction). The third group was told
only that their responses to cold were being
tested (neutral instruction). People who indicated
high anxiety about pain on a questionnaire before
the experiment had the strongest responses — as
measured by the time they kept their hands in
the cold water — not only to the nocebo
instruction, but also to the placebo instruction.
Anyone who is anxious, depressed, or hypochondriacal
runs the risk of developing further symptoms
in response to attempts at healing or comforting.
In this case, the nocebo effect is related to
somatization, the expression of emotional disturbances
in the form of physical symptoms. Somatoform
disorders, identified by recurrent medically
unexplained physical complaints, have many sources
in mood, personality, and social circumstances.
Somatoform reactions may also be provoked and
perpetuated by what some see as the advantages
of being treated as an invalid. This so-called
secondary gain is sometimes regarded as another
form of nocebo response.
Patients need help in understanding and tolerating,
minimizing, or ignoring nocebo and other somatoform
responses. These responses may be at work whenever
the side effects of a medication or other treatment
are vague and ambiguous or the patient has been
expecting it to cause problems. Patients can
be asked about earlier disappointing experiences
with medical procedures. If a patient says he
or she is especially sensitive to drugs, the
physician might point out that anticipating bad
effects can be a self-fulfilling prophecy. It
may help to emphasize the limits of medicine
and explain the close relationship between emotions
and physical sensations, especially as it involves
stress hormones. Above all, in prescribing any
drug or other treatment, physicians must act
in a way that establishes trust and promotes
the patient’s participation and cooperation.
Barsky AJ, et al. “Nonspecific
Medication Side Effects and the Nocebo Phenomenon,” Journal
of the American Medical Association (Feb.
2002): Vol. 287, No. 5, pp. 622–27.
Benedetti F, et al. “Conscious
Expectation and Unconscious Conditioning in Analgesic,
Motor, and Hormonal Placebo/Nocebo Responses,” Journal
of Neuroscience ( May 15, 2003): Vol. 23,
No. 10, pp. 4315–23.
Hahn RA. “The Nocebo
Phenomenon: The Concept, Evidence, and Implications
for Public Health,” Preventive Medicine (Sept.-Oct.
1997): Vol. 26, No. 5, pp. 607–11.
Spiegel H. “Nocebo: The
Power of Suggestibility,” Preventive
Medicine (Sept.-Oct. 1997): Vol. 26, No.
5, pp. 616–21.